A still greater disadvantage of the test in question is 

 that the dry-residue cipher of one and the same kind of rice 

 shows a considerable amount of instability, proportionate to 

 the duration of storage. Rice with its silver-skin is especially 

 subject to this instability, while this sometimes does not 

 appear in completely polished kinds of rice, or at any rate 

 it is not at all visible, or only in a very slight degree, 

 during the first five months. The principle change consists 

 of a gradual decrease in the dry-residue cipher about the 

 third or fourth month of storage. This is to be attributed to 

 physical modifications thereby undergone by the substances 

 which are soluble in spirit. 



Besides this reduction in the dry-residue cipher there is 

 a temporary increase, though in much less degree, and only 

 in the case of unpolished rice. This begins a few weeks 

 after husking, and reaches its toppoint in the second or 

 sometimes in the third month of storage. This is readily 

 explained by the forming of microscopical cracks in the 

 outside silver-skin layer, so that also the inner parts are 

 exposed to extraction by the spirit. 



The progress and the degree of these undesirable changes 

 are illustrated by the followiug table X. With regard to 

 the want of regularity which is shown, the remark ought 

 to be previously made that this is due in great measure, if 

 not entirely, to two circumstances. In the first place, the 

 desired chronological precision could not always be kept in 

 view during the periodical investigations and in the second 

 place, the given kinds of rice, with a single exception, were 

 not personally collected, and there was therefore no absolute 

 certainty that they had actually all been received shortly 

 after husking. Although for these reasons the accuracy of 

 the ciphers given leaves much to be desired, they are 

 however mentioned just as they were found, because in any 

 case the disadvantage of this changeability for the test is 

 thereby clearly demonstrated. 



